Search papers, labs, and topics across Lattice.
3
0
6
Pathology MLLMs can now better incorporate diagnostic standards during reasoning, thanks to a new memory architecture inspired by how human pathologists process information.
Multimodal jailbreaks, meet your match: SaFeR-ToolKit's virtual tool-calling protocol boosts VL model safety by up to 55% without sacrificing general capabilities.
MLLMs encode conflicting knowledge signals as linearly separable features in mid-to-late layers, revealing a distinct processing stage for conflict resolution.
